Anthropic alone is predicted to have $70B-$100B Annual Recurring Revenue by end of 2026. It may not turn out to be the long-term behemoth some are touting, it's way more than a $50B/year industry even today.

Are you implying these folks didn't invent RADAR?

Amazon aside, because I'm not really sure the courts have decided the legality of scanning books for AI training, there's another crime here.

Placing a tracker on an item you sell without informing the buyer is illegal. At least in most of the U.S. The "investigators" in this story likely committed a crime.

Why would they relocate the staff?

Depends on how you look at it. There were more "big" companies in the past, but they've been weakened to the point that consolidation has reduced that to a "few" big companies with lower margins. Not sure if the pool itself is smaller or poorer, but there are definitely less pieces of the pie.

We should tariff this in to oblivion. We need to be making battery cells (of all kinds) domestically. That means we need to adjust regulations to allow it, and also prevent foreign subsidies from pricing competition too low for it to make sense. The latter is literally what tariffs are for.

More laws against killing isn't going to solve the killing problem.

True, which is why they build Collosus near Memphis where there was surplus power available, and for which they applied for and purchased usage of. The remaining power requirements are being provided by NatGas portable (truck) generation until the already applied for permanent power production can be approved and built.

I'm not saying that electrical providers shouldn't require forecasting and approval of power needs of datacenters prior to demand. Of course they should, as should the builder of such datacenters be doing their planning up front also. That's where this argument falls apart...both are happening.

Requires service providers block illegal operations. Why is this "news"?

Whether "AI" is hype or will burst its bubble remains to be seen...nobody can be sure yet, but there are huge bets that it will persist in one form or another.

But your above comments are completely lacking any grounding in facts. At best they are something you read in some FUD newspaper article with an agenda against datacenters for some reason they can't really explain. I'll buy an argument that they make noise...chillers definitely do...but the rest is nonsense and factually untrue.

You could just use DDG or Google Search to learn how to do those things. You could try Bing, but it wouldn't give you any relevant results. I don't see how this part of "AI Safety" is really a thing we should worry about. Rogue autonomous cyber attacks? Sure. Censoring information? Nahh.

This was discussed on the All-In podcast last week. The "temp turbines" are mobile units (truck-based,) and thus classified as temporary consumer use. They don't require permits, which is precisely why Musk's companies acquired controlling stake the company. The permitting process for a permanently installed base of power didn't meet the timeline, but those permit applications are in progress.

I'm not saying that xAI/Musk/whoever you want to blame is going out of their way to protect the environment from emissions, but the insinuation that they are somehow breaking the law/regulations by not having proper permits is patently false and also misleading by indicating intentional wrong-doing.

Isn't that basically what a Chromebook is?

Until someone can really, and I mean REALLY, explain to me what constitutes a token in a way that I can calculate and predict, these sorts of billing models are meaningless. Only then can we evaluate what the costs, risks and benefits of AI use really are in any sort of quantifiable way.